You are on page 1of 3

steelwise

May 2008

Your connection to
ideas + answers

What Every Fabricator Wants


You to Know about Welding
By Monica Stockmann and Thomas J. Schlafly

There are plenty of things that engineers can do to


make the welders work easier and more efficient.

If you asked the opinion of a typical steel


fabricator, he or she would probably tell you that
there should be a class (or two) in engineering and
architecture curriculums that focuses on welding
and connection designfrom both the engineers
and the fabricators perspective. However, theres
simply no space in the typical curriculum for such a
course. To help fill that need, weve developed this
article to cover several welding design recommendations from the fabricators standpoint, including
filler metal specifications, preferred weld types, and
other applicable weld advice. Obviously we cant
cover everything here, so refer to AISCs Design
Guide 21, Welded ConnectionsA Primer for Engineers; the American Welding Society 2006 Structural Welding CodeSteel (AWS D1.1:2006); and
Section J2 of the AISC 2005 Specification for Structural Steel Buildings for additional information.
When performed properly, welding is an economical and efficient tool for joining steel. However, welding can become altogether uneconomical when improperly specified. The skilled labor
cost required to make a weld typically accounts
for 75% to 95% of the total cost of a weld. Thus,
cost-effective welding is typically achieved when
the required weld metal is deposited in the least
amount of time. Here are some tips and highlights
on how to design and specify economic welds.

Welding Processes
In total, there are approximately 100 different
welding processes. At present, the steel construction industry uses about five of them: shielded
metal arc welding (SMAW); flux cored arc welding (FCAW), which can be gas or self shielded;
submerged arc welding (SAW); gas metal arc
welding (GMAW); and electroslag welding
(ESW). Design Guide 21 describes many of the
pros and cons of the different types of welding,
as well as several process-specific welding issues,
such as applicability of low-hydrogen electrodes
in SMAW; the conditions under which short-circuit transfer may occur in GMAW; and the use of
active fluxes in SAW.


The choice of welding process can significantly


affect the cost of a project. Typically, the contractor makes this important decision, as he or she is
usually best positioned to select the optimal welding process for a given application. When properly
used, all of the welding processes listed in AWS
D1.1 are capable of delivering adequate welds for
building construction.
Over-use of CJP welds
It is common knowledge that complete joint
penetration (CJP) groove welds are typically
the most expensive type of weld and thus should
be reserved for situations in which they are the
only viable option. In a CJP groove weld, the full
strength of the attached materials is developed.
Thus, no design calculations are required when
specifying CJP groove welds in statically loaded
structures. However, this design advantage can
be abused by specifying CJP groove welds in situations where there are better, more economical
options. The most commonly abused situation
involves longitudinal welds on built-up beam and
column sections. These welds are typically loaded
in shear, which rarely requires the strength of CJP
groove welds. Fillet welds or PJP groove welds are
typically better, lower-cost options for this case.
Fillet and PJP welds are also typically the more
environmentally friendly option, because depositing less metal saves energy too!
Fillet Welds vs. PJP Groove Welds
A helpful rule of thumb is to use fillet welds
whenever the required weld leg size is 1 in. or less,
and PJP groove welds or a PJP and fillet weld combination when larger sizes are required. Because
most structural steel fillet welds are not required
to have leg sizes greater than 1 in., fillet welds
are typically the most economical choice, with
the exception of skewed T-joints, which must be
examined on a case-by-case basis. The table on the
next page, taken from Design Guide 21, guides the
user through an economical weld selection process
based on loading and joint type.

Monica Stockmann is a Steel


Solutions Center advisor, and
Thomas J. Schlafly is director
of research, both with AISC.

may 2008 MODERN STEEL CONSTRUCTION

Taking the Short CutSpecifying the Joint Strength


In many cases, an alternative to specifying information in the
weld symbol is to specify the required strength of the joint required.
The fabricator will then design the weld to develop the strength
specified as he or she sees economically fit, while still abiding by
the design rules of AWS D1.1. For example, a weld symbol without
a dimension and without CJP in its tail designates a weld that will
develop the adjacent base metal strength in tension and shear.

Table 32. Weld


Selection Based on Loading and Joint Type
4ABLEn7ELD3ELECTION"ASEDON,OADINGAND*OINT4YPE
*OINT4YPE

&ORCE4YPE

4ENSION

"UTT*OINTS

#OMPRESSION

3HEAR

,OADING,EVEL

.ORMALTOTHE
7ELD!XIS

0ARALLELTOTHE
7ELD!XIS

,IGHT

0*0

0*0

(EAVY

#*0

0*0

,IGHT

0*0

0*0

(EAVY

0*0WITHBEARING
CONSIDERED #*0

0*0

,IGHT

0*0

(EAVY

#*0

,IGHT

To return or not to return?


End returns, otherwise known as boxing, are the continuation
of a fillet weld around the corner of a member as an extension of
the primary weld. End returns are used to ensure quality terminations to welds and to provide some resistance to prying of the weld
roots. In general, end returns are neither prohibited nor required.
AISC specification Section J2.2b provides further discussion on the
requirements and limitations for end returns. In statically loaded
structures, fillet welds can be stopped short of the end of the joint by
a length equal to the leg size of the weld, or continue to the end or
be returned around the corner, except as noted in J2.2b(1)-(4). One
exception to note is that for flexible connections, such as framing
angles and tees, the tension edges of the outstanding legs or flanges
must be left unwelded over a portion of their length to assure flexibility of the connection. If end returns are used in this case, their
length must be restricted to not more than four times the weld size or
half the width of the angle, as shown in the following figure.

4ENSION

4EE*OINTS

#OMPRESSION

(EAVY

&ILLET
&ILLET 0*0
0*0&ILLET #*0

&ILLET
&ILLET

,IGHT

&ILLET

&ILLET

(EAVY

0*0WITHBEARING
CONSIDERED #*0

&ILLET

,IGHT
3HEAR

4ENSION
#ORNER
*OINTS
/UTSIDE

#OMPRESSION

3HEAR

#ORNER
*OINTS
)NSIDE

#OMPRESSION

&ILLET
&ILLET 0*0
0*0&ILLET #*0

(EAVY
,IGHT

0*0

0*0

(EAVY

#*0

0*0

,IGHT

0*0

0*0

(EAVY

0*0WITHBEARING
CONSIDERED #*0

0*0

,IGHT

0*0

(EAVY

#*0

,IGHT
4ENSION

(EAVY

&ILLET
&ILLET 0*0
0*0&ILLET #*0

&ILLET
&ILLET

,IGHT

&ILLET

&ILLET

(EAVY

0*0WITHBEARING
CONSIDERED #*0

&ILLET

,IGHT
3HEAR

,AP*OINTS

3HEAR

3HEAR

(EAVY

&ILLET
&ILLET 0*0
0*0&ILLET #*0

,IGHT

&ILLET 0LUG3LOT

(EAVY

&ILLET 0LUG3LOT
&ILLET0LUG3LOT

Source: AISC Design Guide 21, Welded ConnectionsA Primer for Engineers

  $%3)'. '5)$%  7%,$%$#/..%#4)/.3!02)-%2&/2%.').%%23

C-shape with vertical line indicates weld is to be returned.

End return length = 2w (preferred), with a maximum of 4w or Lo,


whichever is less.

There is no defined or code-specified standard symbol for


welded clip angle connections. Some firms include figures and
descriptions of their typical weld practices on the general detail
sheet of their shop drawings. Thus, it would be reasonable to
include typical weld end condition details on this sheet as well.
Some commonly used weld symbols for a returned and a stoppedshort (non-returned) weld are shown in the figures below. In addition, a commonly used symbol for an angle that is welded on three
sides is also shown below. These symbols are not AWS standards;
they are simply suggestions.

MODERN STEEL CONSTRUCTION may 2008

C-shape indicates weld is to be made on three sides of the angle, full


length.

The pair of vertical lines, one shorter than the other, indicates weld is
to be stopped short.

Why Weld All Around?


The weld all around symbol specified by the designer is a
convenience that often results in unnecessary work and confusion
for the fabricator. The symbol is all too often used in applications
where the weld is not necessarily required on the entire perimeter
of a connection, such as certain types of column-to-base plate connections. Instruction for the fabricator to weld all around typically
requires him to reposition the steel pieces several times, and the
process often requires welding in areas that are difficult to access.
To avoid specifying a fabrication headache, the engineer should
think twice about how much weld is actually needed in the connec-

tion. For a column-to-base plate connection, perhaps a weld along the top of one
flange and the bottom of the other flange
would sufficeand the fabricator will be
able to access the specified weld areas easily, even without having to reposition the
members.
CVN Toughness
Toughness is a material property that
measures the resistance to cracking in certain conditions. Charpy V-Notch (CVN)
toughness is a test requirement related to
material toughness. Some filler metal classifications include CVN requirements. In
the U.S., common classifications require
a CVN toughness of 20 ft-lb at -20F or
0F as used in seismic applications. If notch
toughness of welded joints is required, the
engineer needs to specify the minimum
absorbed energy and the corresponding
test temperature for the filler metal classification to be used. As an alternative, the
engineer needs to specify that the Weld
Procedure Specification (WPS) be qualified with CVN tests.
Weld Design Information
Engineers often provide guidance on

welding in design drawings and specifications in order to prevent misunderstandings and to emphasize code requirements.
Although the design professional may
intend to avoid past welding conflicts or
errors, the welding information provided
on their design drawings can often conflict
with code requirements and cause unintended confusion.
Most welds performed on structural
steel are meant to be conducted in accordance with prequalified weld procedures.
The code includes provisions for selection
of appropriate welding processes, filler
metal classifications, joint designs, procedure variables, and other provisions guiding
fabrication. Weld procedures that are not
prequalified have to be tested, and the AWS
D1.1 limits the essential variables to ranges
around those that were tested. Information
about these items on design drawings may
not be helpful, and in some cases may conflict with good practice. For example, do
you really want to specify E70 in cases that
AWS D1.1 would indicate an E80 electrode
for weathering characteristics?
Chapters 1 and 2 of AWS D1.1 illustrate
all of the specific welding requirements that

the engineer shall specify in the contract


documents as necessary. Some key things to
remember include: tell the fabricator what
strength you assumed when sizing fillet and
pipe welds; tell the fabricator to comply
with AISC and AWS; and then check Chapters 1 and 2 of the AWS code to see if you
need to tell the fabricator any more.
Call to Action
Proper weld design and specification
results in significant project cost savings.
Although many engineers enter the workforce with little or no weld design experience, it is important for them to research,
learn, and master welded connection
design based on both safety and economy.
The best weld detail for a specific connection is one that reliably and safely transmits
the imposed loads, and yet is economical
and easily made by the welder. Your challenge: put these lessons learned to practice
in your everyday design work, and you will
surely put a smile on the welders face during fabrication.

may 2008 MODERN STEEL CONSTRUCTION